Tuesday, July 21, 2015

Review of the Drunkard's walk by Mlodinow: To increase your number of successes you must simply increase your number of failures

This was a great book. It had just the right balance of anecdotes, mathematics, scientific studies and history to offer the reader a comprehensive and informative, yet thoroughly enjoyable introduction to the field of randomness. As the author rightly points out, again and again, people are blind when it comes to the role that chance or randomness plays in their lives, which is in fact very big. We tend construct our life narrative around situations where we made a decision that seemed to be crucial, which makes it seem as if we have been in the driving seat for much of our life. Still, most people can come up with seemingly random events that shaped the rest of their life. For example, I would never have met my wife had I not turned down a job one summer 10 years ago and I would not have ended up as a scientist had my grades been good enough to become a psychologist. If you would change just a few minor details in my history, and I might have lived a completely different life today. 

Mlodinow begins the book by discussing some real life examples where people often fail to see the underlying mathematical truths. When a company does well, a CEO is rewarded with sometimes ridiculous bonuses, only to be fired the next year because the company suddenly did fare so well. This is the case despite the fact that fluctuations in the market are inescapable. The same is true in the world of sports where managers are frequently fired following dips in form which necessarily occurs if luck is a factor which it always is in sports.

After this introduction Mlodinow goes through the history of probability theory. I was surprised to learn that the Greek really didn’t get probability. They were excellent when it came to mathematical axioms and deducting knowledge, however, they apparently thought uncertainty had no place in maths and therefore ignored the field entirely. More than a thousand years passed before a man began to investigate the rules of probability in the mid 16th century. His name was Cardano and he was, of course, a gambler. With some very elementary knowledge regarding uncertainty, Cardano won lots of money which he used to finance his studies in Medicine.

Mlodinow continuous to move through history, while also making sure that the reader understands the theories that are being developed. Among others one encounters Galileo, Pascal, Bernoulli and Laplace who all worked on probability in different ways. One learns about the normal curve, chaos theory and bayesian statistics. Again, everything is written in an engaging yet simple fashion and I personally felt I learnt a lot even though I have studied statistics at University.

This book also deserves credit for being the first to explain the Monty Hall problem in a way that made me feel I really get it. Imagine you are a contestant on TV show, there are three doors and behind one of them is a car, while the other two doors have goats behind them. You pick one door (that you don’t open), then the TV host open one of the other two doors behind which there is a goat. At this point you have to choose to open the door you picked initially, or switch to the other door. What do you do? Even though more than 90% in polls, as well as thousands of mathematicians, passionately believed that it did not matter whether or not you switched, the correct answer is in fact that you will double your chance of winning if you switch to the other door. As Mlodinow explains you really have to guess which of the following two scenarios you are in:

1. The door you initially picked was the correct one (chance one third). If you switch you will find a goat.

2. You initially picked the wrong door (chance two thirds). Since the host will always open a door with a goat the correct one is the one the host did not open and which you did not pick. If you switch you will win.

In other words, if you picked the wrong door initially you will win if you switch and since it is more likely to pick the wrong door than the right door your chances are better if you switch.

In the last part of the book, Mlodinow return to the role that randomness plays in our life. After he has convincingly demonstrated how great this role is he arrives to the question of how one should act in the face of such uncertainty. Given that our successes or failures to a large extent are a result of random events, should we just stop trying? No! Mlodinow eventually arrives at the quote that is the title of my review. If you want to increase your success rate, you should increase your failure rate. Those who succeed in the end tend to be those who try again and again and again i.e. those who throw the dice over and over again will, eventually, end up with a six. Having read this book, I am determined to go out in the world and start failing. Thank you Mlodinow for the inspiration and for this excellent book!

Tuesday, March 24, 2015

How to create a detailed false memory

Courts today rely to a large extent on eyewitness testimony, both from the person who has been accused as well as from bystanders who happened to be there at the time of the crime. Indeed, if a person confesses to a crime then he is almost certain to be convicted. All this assumes that eyewitness testimonies are generally accurate descriptions of past events, however, there are by now many many studies showing that it is easy to make people believe in things that did not happen. One of my favorite (and funny) examples is a study where participants were led to believe that they had tea with Prince Charles. However, there are also a number of not so funny examples of false memories, such as the memories that led to conviction of many innocent people on charges of child sexual abuse or ritual satanic murders or both.

In a new study published in Psychological Science the authors wanted to create false memories that were more detailed and contained multisensory details, such as smells and feelings, than in previous studies. Most people tend to have more confidence in memories where if they remember how they felt or if they remember a certain smell, than in memories that lack such details.

To instill the false memories each of the 60 participants, aged 18-31, went through three different interviews. In the interviews the naive participants were told about two different events that occurred when they were between 11-14 years old. One of the events had actually happened while the other was made up. The aim of the study was to see if the participants could be led to believe that the false event had actually happened, and to compare the memory of these false events with the memory of the events that had really happened. The false memories were far from mundane and involved criminal acts such as assault. The interviewer used a number of techniques that are common practice for inducing false memories such as…
  • Using incontrovertible evidence (“In the questionnaire, your parents/caregivers said...”)
  • Social pressure (“Most people are able to retrieve lost memories if they try hard enough”
  • Building a good relationship by asking how the students semester, nodding and smiling a lot
  • Using long pauses or asking “what else” to encourage the student to provide additional details. 
The results in this study showed that even with a very conservative measure, 44 of the 60 students had acquired a false memory of their own, serious crime, after the three interviews (in the end they were of course told that the false event had not actually happened). The criteria for saying that they had adopted the false memory included that they said they believed that it had happened and that they had provided a number of details concerning the event beyond what the interviewer had told them. With less strict criteria (e.g. just saying that they had believed that the event had occurred), 54 of the 60 student had adopted the memory. In other words, of the 60 student, only 6 appeared to be at least somewhat immune to the false memory.

Those students who adopted the false memory could provide detailed descriptions of the events such as how anxious they felt. They were also able to describe people who “had been there”, and they reported that they thought about the event outside of the interview situation. In other words, these false memories were rich in details and courts would have had no reason to doubt the validity of these memories. An important lesson from this study is therefore that we cannot trust a memory just because we think we remember the context. As a family man I am also struck by how often it happens that peoples memory contradict each other. “But I told you that you should by xxx”, or “We agreed that xxx”. Clearly, false memories arise outside the laboratory

In conclusion, once again, a study has revealed how error-prone our memory really is. Personally I try to be extremely skeptical with regards to my own memory and I try to never make any certain assertions based on “what I remember”. I think that the court system should probably tune up their skepticism towards eyewitness testimonies a few steps, unless they want to keep sending innocent people to jail.

Shaw J, & Porter S (2015). Constructing Rich False Memories of Committing Crime. Psychological science PMID: 25589599